Decentralized smoothing algorithms of continuous-time systems in distributed-sensor networks

Some decentralized smoothing algorithms are derived by applying a Rauch-Tung-Striebel fixed-interval smoother formula in continuous-time systems supposing that an estimation structure is comprised of a global processor and of two local processors. Two cases are investigated for the problems of decentralized smoothing: when the local-filtered estimated are available and when the local-smoothed estimates are known. Some features of present algorithms are discussed from the point of view of data transmissions and communication bandwidth, etc.     

分布网络中离散时间系统的平滑算法*

本文考虑具有一个中心处理器和两个局部处理器下的估计结构的平滑合成问题,讨论了当已知局部滤波估计和当已知局部平滑估计时两种情况下的分布平滑估计,并导出了合成算法。     

A scattering framework for decentralized estimation problems

In this paper we develop a comprehensive framework for the study of decentralized estimation problems. This approach imbeds a decentralized estimation problem into an equivalent scattering problem, and makes use of the super-position principle to relate local and centralized estimates. Some decentralized filtering and smoothing algorithms are obtained for a simple estimation structure consisting of a central processor and of two local processors. The case when the local processors exchange some information is considered, as well as the case when the local state-space models differ from the central model.     

Continuous-time decentralized smoothers based on two-filter form: identical local and global models

This paper proposes some decentralized smoothing algorithms for a continuous-time linear estimation structure consisting of a central processor and of two local processors, in which the local models are assumed to be identical to the global model. The philosophy of the paper is to solve the problem in terms of the local forward and backward information (or Kalman) filters. The resulting algorithms are somewhat different from those based on the local smoothing estimates which have been studied by some other authors. Smoothing update and real-time smoothing algorithms are also presented, ft is shown that the present algorithms have some advantages: the global filtered estimates can be obtained in the course of computing the decentralized smoothing estimates and the central and local processors can be derived in a completely parallel fashion     

Fixed-point smoothing with a measurement process initiated a posteriori

For discrete linear stochastic systems, a fixed-point smoothing algorithm is given to obtain optimal estimates of the state of the system at timek = i(ifixed) when noisy linear measurements of the state are not available until the timek = j(j > i). An example is included to illustrate smoothing degradation as (j - i) increases.     

Forward/Backward contour tracing with feedback

This correspondence describes a contour extraction algorithm which can gradually improve its results until the extracted contours are closed. This is achieved by an architecture with a feedback path for local smoothing. The feedback path is activated only when one or more contours obtained are not closed in order to initiate smoothing in noisy areas of the image to remove local irregularities that cause the problems. A forward/backward boundary tracing mechanism is employed to facilitate locating any troubled areas. A smoothing method appropriate for reducing local irregularities is discussed. The proposed algorithm is very suitable for those applications that demand closed contours, such as character recognition and blob detection.     

Graph-theoretic approach to controllability and localizability of decentralized control systems

The controllability and the localizability problems are considered under the decentralized information structure using some concepts from graph theory. First of all, the information structure graph is introduced for a decentralized control system based upon local output controllability, matrices between stations when a local output feedback law is applied, and some invariant properties of the graph with respect to local output feedback laws are driven. Then by using these results, necessary and sufficient conditions of controllability under decentralized information structure are obtained. Also obtained are several conditions for localizability, of decentralized control systems, which is newly introduced in the present paper as an extended concept of a class of disturbance localization and decoupling for centralized systems.     

广义分散前馈控制系统固定模的进一步研究

运用把广义分散前馈控制系统转化为不带前馈的广义分散控制系统的方法,讨论了广义 分散前馈控制系统的固定模问题,给出了若干以系统原始矩阵为基础的有穷固定模和脉冲固 定模的新判据,得到了一些有意义的结论.     

广义分散前馈控制系统固定模的进一步研究

运用把广义分散前馈控制系统转化为不带前馈的广义分散控制系统的方法，讨论了广义分散前馈控制系统的固定模问题，给出了若干以系统原始矩阵为基础的有穷固定模和脉冲固定模的新判据，得到了一些有意义的结论．     

Decentralized filtering adaptive constrained tracking control for interconnected nonlinear systems

This paper presents a novel decentralized filtering adaptive constrained tracking control framework for uncertain interconnected nonlinear systems. Each subsystem has its own decentralized controller based on the established decentralized state predictor. For each subsystem, a piecewise constant adaptive law will generate total uncertainty estimates by solving the error dynamics between the host system and decentralized state predictor with the neglection of unknowns, whereas a decentralized filtering control law is designed to compensate both local and mismatched uncertainties from other subsystems, as well as achieve the local objective tracking of the host system. The achievement of global objective depends on the achievement of local objective for each subsystem. In the control scheme, the nonlinear uncertainties are compensated for within the bandwidth of low‐pass filters, while the trade‐off between tracking and constraints violation avoidance is formulated as a numerical constrained optimization problem which is solved periodically. Priority is given to constraints violation avoidance at the cost of deteriorated tracking performance. The uniform performance bounds are derived for the system states and control inputs as compared to the corresponding signals of a bounded closed‐loop reference system, which assumes partial cancelation of uncertainties within the bandwidth of the control signal. Compared with model predictive control (MPC) and unconstrained controller, the proposed control architecture is capable of solving the tracking control problems for interconnected nonlinear systems subject to constraints and uncertainties.     

Stabilizing bit-rates in networked control systems with decentralized event-triggered communication

In a decentralized event-triggered networked control system (NCS), an agent samples and transmits its local state information to the controller when some local event occurs. Such event-triggered NCSs were expected to be more efficient than traditional periodically sampled system in terms of communication channel usage. This paper studies the stability of decentralized event-triggered NCS in the presence of quantization and delays. We point out some potential issues in decentralized event-triggered design and propose an alternative decentralized event with a linear-affine threshold, which avoids infinitely fast data transmission. Conditions on quantizer and communication channel are derived, which, when satisfied, can guarantee stability of the resulting NCS. Based on these conditions, finite stabilizing bit-rates are provided.     

广义分散控制系统脉冲固定模的代数特征

本文进一步讨论了分散脉冲固定模的代数特征，得到了若干判别分散脉冲固定模，分散输入，输出脉冲固定模以及分散正常化的充要条件。     

高斯混合概率假设密度无序估计分布式融合

针对分布式传感器网络中多目标随机集状态混合无序估计问题,本文提出了一种基于高斯混合概率假设密度无序估计分布式融合算法.在高斯混合概率假设密度滤波器的框架下,首先基于概率假设密度递推滤波特性,建立适用于多目标随机集状态混合无序估计的最新可利用估计判别机制,然后利用扩展协方差交叉融合算法对经过最新可利用估计判别机制获得的无序概率假设密度强度估计进行融合处理,针对融合过程中高斯分量快速增长的问题,在保证信息损失最小的前提下,对融合过程的不同环节实施高斯混合分量裁剪操作,给出了一种多级分层分量裁剪算法.最后,仿真实验验证了文中所提的算法的有效性和可行性.     

非凸体分步逐次高阶磨光方法研究

基于KS函数的包络特性和内点极径扫描法,提出了一种新的非凸体高阶磨光方法——分步逐次高阶磨光方法．此方法将磨光过程分为两步：第一步对每一局部非凸部位依次进行高阶磨光,第二步对剩余的局部凸域以及经第一步磨光后所得的非凸域进行整体包络高阶磨光．典型实例表明,这种新的磨光方法对解决某些非凸体高阶磨光问题具有一定的优越性。     

Two-level stochastic LQ control problem for a dynamic system

Two-level, hierarchical, partially decentralized control for a large-scale system is considered. It is seen that for an LQ problem the optimal control laws are linear functions of the state estimates. The filter equations for some state estimates, determined by a coordinator and local controllers co-operating in a two-level control structure, are derived, A discussion on the usefulness of the application of particular local controllers is given, as well as a numerical example.     

Online planning for multi-agent systems with bounded communication

We propose an online algorithm for planning under uncertainty in multi-agent settings modeled as DEC-POMDPs. The algorithm helps overcome the high computational complexity of solving such problems offline. The key challenges in decentralized operation are to maintain coordinated behavior with little or no communication and, when communication is allowed, to optimize value with minimal communication. The algorithm addresses these challenges by generating identical conditional plans based on common knowledge and communicating only when history inconsistency is detected, allowing communication to be postponed when necessary. To be suitable for online operation, the algorithm computes good local policies using a new and fast local search method implemented using linear programming. Moreover, it bounds the amount of memory used at each step and can be applied to problems with arbitrary horizons. The experimental results confirm that the algorithm can solve problems that are too large for the best existing offline planning algorithms and it outperforms the best online method, producing much higher value with much less communication in most cases. The algorithm also proves to be effective when the communication channel is imperfect (periodically unavailable). These results contribute to the scalability of decision-theoretic planning in multi-agent settings.     

Decentralized control: a stable proper fractional approach

Three decentralized control problems are considered in a fractional setup for two-channel multivariable systems. All three problems are instances of decentralized control or local output feedback problems. The problems are: (i) making the system stabilizable and detectable through the first channel via dynamic output feedback around the second channel, (ii) the first problem with the constraint of internal stability, and (iii) the decentralized stabilization problem. All three problems are equivalent as far as the solvability conditions are concerned. A characterization of all solutions in each case is given. The results apply to a class of systems having fractional representations over an arbitrary principal ideal domain     

Robustness of decentralized knowledge systems: a grammar-theoretic view

This paper introduces the formal framework of grammar systems to handle a practical and important property of decentralized rule-based knowledge systems. The property is called robustness. In our framework, a rule-based system is robust when some rules can be removed from it and yet its critical functionality remains unchanged. As a theoretical framework for study robustness of decentralized rule-based systems we use grammar systems. We prove within that framework that the question whether a knowledge system is robust or not is undecidable. In contrast, we prove with the same framework that whether or not a component is ever enabled, or whether or not a component working in a special—so called maximal—mode blocks the further functioningof a systems when enabled, are decidable. Some open problems are also formulated.     

Nonparametric Approximation of the Density Function Logarithm of the Two-dimensional Random Variable

Consideration was given to a method for approximating the probability density of a two-dimensional random variable with separate stages of generating the local estimates and smoothing the random errors. It was proposed to decompose the space on the basis of the source data into minimal-size domains where the local estimates of the density logarithm which correspond to the observation model with additive errors accepted in the regression analysis were generated. The error covariance matrix was shown to be completely definite and independent of the original density, which made it possible to apply the apparatus of nonparametric regression to optimizing the choice of the smoothing parameter.     

Centralized and decentralized asynchronous optimization ofstochastic discrete-event systems

We propose and analyze centralized and decentralized asynchronous control structures for the parametric optimization of stochastic discrete-event systems (DES) consisting of K distributed components. We use a stochastic approximation type of optimization scheme driven by gradient estimates of a global performance measure with respect to local control parameters. The estimates are obtained in distributed and asynchronous fashion at the K components based on local state information only. We identify two verifiable conditions for the estimators and show that if they, and some additional technical conditions, are satisfied, our centralized optimization schemes, as well as the fully decentralized asynchronous one we propose, all converge to a global optimum in a weak sense. All schemes have the additional property of using the entire state history, not just the part included in the interval since the last control update; thus, no system data are wasted. We include an application of our approach to a well-known stochastic scheduling problem and show explicit numerical results using some recently developed gradient estimators